Full Terms & Conditions of access and use can be found athttp://www.tandfonline.com/action/journalInformation?journalCode=ueso20

Download by: [Huazhong University of Science & Technology ] Date: 07 December 2016, At: 02:41

Energy Sources, Part A: Recovery, Utilization, andEnvironmental Effects

ISSN: 1556-7036 (Print) 1556-7230 (Online) Journal homepage: http://www.tandfonline.com/loi/ueso20

Utilization of pyrolytic char derived from bamboochips for furfural removal: Kinetics, isotherm, andthermodynamics

Y. C. Li, X. H. Wang, J. A. Shao, P. Li, H. P. Yang & H. P. Chen

To cite this article: Y. C. Li, X. H. Wang, J. A. Shao, P. Li, H. P. Yang & H. P. Chen (2016) Utilizationof pyrolytic char derived from bamboo chips for furfural removal: Kinetics, isotherm, andthermodynamics, Energy Sources, Part A: Recovery, Utilization, and Environmental Effects,38:11, 1520-1529, DOI: 10.1080/15567036.2014.940093

To link to this article: http://dx.doi.org/10.1080/15567036.2014.940093

Published online: 26 May 2016.

Submit your article to this journal

Article views: 52

View related articles

View Crossmark data

Utilization of pyrolytic char derived from bamboo chips for furfuralremoval: Kinetics, isotherm, and thermodynamicsY. C. Li, X. H. Wang, J. A. Shao , P. Li, H. P. Yang, and H. P. Chen

State Key Laboratory of Coal Combustion, Huazhong University of Science and Technology, Wuhan, China

ABSTRACTBamboo charcoal obtained from the pyrolysis of bamboo chips was used toremove furfural, a representative fermentation inhibitor in hydrolyzates orpyrolysis oil. Kinetics, isotherm, and thermodynamic calculations were per-formed. The adsorption process can be well depicted by Ho’s pseudo-second-order model. The particle diffusion of homogeneous particle diffu-sion model (HPDM) and shrinking core model (SCM) was found to be thecontrolling step. Isotherm analysis indicated the adsorption feature tookplace by a nonideal adsorption. Thermodynamic calculation suggested thatΔH° > 0, ΔG° < 0, ΔS° > 0, and Ea value was 2.02 kJ/mol. The comparisonresults demonstrate bamboo charcoal is a promising adsorbent for furfuralremoval.

KEYWORDSAdsorption kinetic; furfural;pyrolytic char;thermodynamic

1. Introduction

Bioethanol, a renewable biofuel, has drawn great attentions nowadays because of the shortage ofpetroleum, the increasing global energy demand, and the large amounts greenhouse gas emission(Balat, 2009). The lignocellulosic biomass can be used as a potential feedstock for bioethanol production(Demirbas, 2006), because it mainly consists of hemicellulose and cellulose. The hemicellulose andcellulose are the structures of polymers of five/six carbon chain sugar (Ranjan et al., 2009). Thesepolysaccharides could be decomposed into monosaccharide by enzymatic/acid hydrolysis or by thepyrolysis process (Li and Zhang, 2004). The monosaccharide can be used as the carbon sources formicrobial fermentation. However, many kinds of inhibitor compounds will be generated in the trans-formation process of biomass, including aldehydes, acids, and so on (Parawira and Tekere, 2011), andfurfural is one of the representative fermentation inhibitors that strongly inhibits the growth of micro-organisms (Xiros et al., 2011). Thus, furfural must be removed first before bioethanol production.

Several attempts have been made for the removal of furfural along with some other fermentationinhibitors from the pyrolysis oil or hydrolyzates, and the adsorption method has proved to be a reliabletechnology (Ranjan et al., 2009). In the authors’ previous study, a cheap pyrolytic char derived frombamboo chips (named bamboo charcoal) was used to remove furfural, and it was found that charcoalcould selectively adsorb furfural despite the presence of sugars due to the π-π dispersion interactionbetween furfural and the bamboo charcoal (Li et al., 2013). However, the adsorption kinetics andisotherm along with the adsorption thermodynamic process of furfural in detail by this pyrolytic charwere still not comprehensive. The kinetics, isotherm, and thermodynamic parameters are very importantfor the adsorption design, as some of themmay provide an insight into understanding the mechanism ofthe adsorption process. Benamor et al. (Benamor et al., 2008) studied the kinetic process of cadmium ionsonto XAD-7 and found the rate-limiting step was particle diffusion. He also quantitatively calculated the

CONTACT Jingai Shao jashao@hust.edu.cn State Key Laboratory of Coal Combustion, Huazhong University of Science andTechnology, 1037 Luoyu Road, Wuhan 430074, China.Color versions of one or more of the figures in the article can be found online at www.tandfonline.com/ueso.© 2016 Taylor & Francis Group, LLC

ENERGY SOURCES, PART A: RECOVERY, UTILIZATION, AND ENVIRONMENTAL EFFECTS2016, VOL. 38, NO. 11, 1520–1529http://dx.doi.org/10.1080/15567036.2014.940093

diffusion coefficient of the adsorption. Hameed and Rahman (Hameed and Rahman, 2008) deduced thatthe adsorption of phenol on activated carbon was a monolayer adsorption based on the fitting result ofthe isotherm. Qiu et al. (Qiu et al., 2012) investigated the thermodynamic phenomenon of the adsorptionof p-xylene onto activated carbon, and found the process was endothermic, spontaneous, and physics.Because of the specific structure of pyrolytic char derived from bamboo chips that was quite differentfrom activated carbon or polymers, the kinetic, isotherm, and thermodynamic mechanisms on thispyrolytic char still need to be studied. Furthermore, the accurate adsorption capacity of furfural onto thischarcoal remains unveiled.

In this paper, the adsorption kinetics and isotherm of bamboo charcoal for the adsorptiveremoval of furfural are investigated. The rate-limiting step is determined by the homogeneousparticle diffusion model (HPDM) as well as the shrinking core model (SCM). Thermodynamicanalysis is also performed. The basic data obtained will be very important for further adsorptiondesign. In addition, the results are compared with those of the commercial activated carbons andother adsorbents to further shed some light on the possibility of using bamboo charcoal as theadsorbent for furfural removal, which may partially replace the commercial activated carbon at arelatively much lower cost and for better waste reutilization.

2. Materials and methods

2.1. Adsorbent and adsorbate

The bamboo charcoal used here was prepared at 823 ± 20 K in a fluidized bed reactor using bamboochips as the feedstock. The physical/chemical properties of the bamboo charcoal have been providedin Ref. (Li et al., 2013). The charcoal of 0.25–0.83 mm particle size was collected for later usage.

The representative fermentation inhibitor solution was prepared by dissolving a precise quantity offurfural (analytical grade; >99% purity) into deionized water. The concentrations of furfural were5–20 g/L, similar to the inhibitor concentration in pyrolysis oil. Furfural was purchased from Jin ShanTing Xin Chemical Ind. Ltd. (Shanghai, China) and it was used without any further purification.

2.2. Adsorption experiments

The adsorption experiments were performed as follows. Furfural solution (100 ml) with known initialconcentration (C = 5–20 g/L) and a known adsorbent dose (m = 75 g/L) was added into a 200 mlconical flask. The mixture was then agitated at 100 rpm in a temperature-controlled (T = 298–328 K)shaker. The bamboo charcoal was then separated from the solution at different time intervals byfiltering. A Lambda 35 UV-spectrophotometer (PerkinElmer, USA) was used to analyze the residualfurfural concentration. For adsorption isotherms, different concentrations (C = 5, 8, 10, 15, 20 g/L) offurfural solution were agitated at a sorbent dose of 75 g/L until the equilibrium was reached. The blankexperiment was also conducted by agitating furfural solution but without adding bamboo charcoals.Preexperiment showed that furfural was stable with time. The repeatability of all the results was greaterthan 95%.

2.3. Kinetic study

The instantaneous behaviors of the “bamboo charcoal-furfural” system were modeled by Lagergren’spseudo-first-order equation and Ho’s pseudo-second-order equation. The linear forms of these twomodels are lg(qe-qt) = lgqe-k1t/2.303 and t/qt = 1/(k2qe

2) + t/qe, respectively (Allen et al., 2005). Here,k1 and k2 stand for the rate constant of Lagergren’s model (min−1) and Ho’s model (g/mg min),respectively; qe and qt stand for the quantity of furfural adsorbed at equilibrium and at time t (mg/g),respectively. The applicability of the kinetic models could be checked by their linearity.

ENERGY SOURCES, PART A: RECOVERY, UTILIZATION, AND ENVIRONMENTAL EFFECTS 1521

2.4. Rate-controlling steps determination

The adsorption of furfural onto bamboo charcoal can be described by three sequential processes:diffusion of furfural through the liquid film, through the particle, and then chemical reaction withbamboo charcoal. One step that offers much larger resistance than the other two is believed to be the rate-limiting step (Dana and Wheelock, 1974). Two main kinetic models (HPDM and SCM) are selected todetermine which mechanism controls the furfural uptake. For HPDM, if the adsorption process iscontrolled by the liquid film, equation of −ln(1−X(t)) = Kli t (labeled as Eq. (1)) can be used (Valderramaet al., 2008); if it is controlled by particle diffusion, −ln(1−X2(t)) = 2Kt (labeled as Eq. (2)) can be used.Here, X(t) is defined as “qt/qe”, and Kli and K are the rate constants for HPDM and SCM, respectively(Valderrama et al., 2010). For SCM, equations of X(t) = 3CAoKmAt/arCso (Eq. (3)), 1−3(1−X(t))

2/3 + 2(1−X(t)) = 6DeCAot/ar

2Cso (Eq. (4)), and 1−(1−X(t))1/3 = KsCAot/r (Eq. (5)) can describe the process if the

adsorption is controlled by liquid film, particle diffusion, and chemical reaction, respectively. Here,De isthe effective diffusion coefficient. The definitions of the other kinetic parameters have been provided inRef. (Valderrama et al., 2010; Valderrama et al., 2008). The target of this kinetic study was to determinethe rate-controlling steps of furfural onto bamboo charcoal and obtain more kinetic parameters forrepresentative fermentation inhibitor removal.

2.5. Adsorption isotherms

The adsorption isotherms are fundamental in the description of the interactive reaction betweenadsorbent and solutes. Two widely used isotherms (Langmuir/Freundlich isotherm) are selected toevaluate the interactive behavior. The Langmuir isotherm usually supposes a monolayer sorption onthe surface of identical sites and the Freundlich isotherm is usually used in a nonideal adsorptionsystem. The linear forms of these two isotherms are Ce/qe = Ce/qm + 1/(qmKL) and lgqe = lgCe/n +lgKF, respectively (Pehlivan and Arslan, 2006). Here, qm is the maximum adsorption capacity offurfural onto bamboo charcoal (mg/g), and n stands for the adsorption intensity. A value of 1/n < 1means a favorable adsorption process.

3. Results and discussion

3.1. Adsorption kinetics for furfural

The evolvement of furfural uptake qt (mg/g) with time at different initial concentrations andtemperatures is presented in Figure 1a and 1b. It can be seen that the equilibrium of adsorption isachieved within 24 h. The quantity of adsorbed furfural increases with the increase of the initialconcentration and temperatures. However, the effect of temperature on the adsorption of furfural isnot very significant compared to the influence of initial concentrations. The kinetic parametersobtained from the fitting of Lagergren’s and Ho’s models are listed in Table 1. It can be seen thatHo’s model agrees with the experimental data well, and is better than Lagergren’s model for thewhole sorption stage based on its high correlation coefficients (R2). The values of qe,cal calculated byHo’s model are much consistent with the experimental data than that of Lagergren’s model. The k2values increase with the increase of temperatures, indicating an increasing adsorptive rate. However,it was not observed with the increase of initial concentrations and a similar trend can be found inother research (Sahu et al., 2008).

3.2. Kinetic mechanisms

To realize the adsorption process from a microscopic point of view, kinetic mechanisms mentioned insection 2.4 were evaluated. The experimental data with 10 g/L of furfural were analyzed using the kineticmodels defined by Eqs. (1)–(5), and the results are shown in Figure 1c, in which F(X) is the function ofX(t)(Valderrama et al., 2008). It can be observed that neither the liquid film diffusion model nor the chemical

1522 Y. C. LI ET AL.

reaction model gives straight lines. In general, if it is controlled by film diffusion, the conditions mainlycomprise small particle size, low solution concentrations, and low degree of agitation (Dana andWheelock,1974). The adsorptionprocess in this studydoes notmeet the point. Simultaneously, the chemical reaction isusually considered to be too quick to influence the whole adsorption process (Saha et al., 2004). Thus filmdiffusion and chemical reaction could be excluded as the controlling step. Serarols (Serarols et al., 2001)investigated the adsorption of Zn(II) onto XAD2, a macroporous resin. They found the dominant control-ling-step of the adsorption is the liquid filmmass transfer at low concentrations while it is the intra-particlediffusion for high concentrations. From Figure 1c it can be seen that both HPDM and SCM controlled byparticle diffusion agree with the data well for the overall process. In order to examine the generalapplicability, the HPDM and SCM models controlled by particle diffusion were further performed to fitthe other adsorption data. The results are shown in Figure 2 for different initial concentrations (5–20 g/L)and temperatures (298–328 K). The linear regression analysis of the parameters is listed in Table 2.

From Figure 2 and Table 2, it is suggested that the particle diffusion control is the dominant transportrate for furfural adsorption onto bamboo charcoal since all of them produce straight lines and have highcorrelation coefficients, specifically at the concentration of 5–20 g/L of furfural at 298 K. Thus, it can bededuced that the furfural removal efficiency will be enhanced by increasing the force of interparticle masstransfer. However, with the increasing temperatures, a slight deviation from linearity was observed usingthese two kinetic models. Kinetic models defined by Eqs. (1), (3), and (5) were also used to fit the data andnone of them gave a straight line (not shown here). That is to say, neither the liquid film diffusion nor thechemical reaction is the rate-limiting step when the adsorption occurs at a high temperature. Shakir and

0 240 480 720 960 1200 14400

20

40

60

80

100

Time/min

5 g/L 10 g/L 20 g/L

(a)

0 240 480 720 960 1200 14400

20

40

60

80

q t/(

mg/

g)

q t/(

mg/

g)

Time/min

298 K 318 K 328 K

(b)

0 120 240 360 480 600 7200.0

0.5

1.0

1.5

2.0 HPDM: Ln(1-X)

HPDM: Ln(1-X2)

SCM: X

SCM: 1-3(1-X)2/3

+2(1-X)

SCM: 1-(1-X)1/3

F(X

)

Time/min

(c)

0 3000 6000 9000 120000

40

80

120

Freundlich• Langmuir

Ce/

q e

Ce/(mg/L)

4.0 3.5 3.0

1.2

1.5

1.8

2.1lgq

e=0.804lgC

e-1.179

R2=0.9986

KF=0.662

Ce/q

e=0.0095C

e+11.5599

R2=0.9898

KL=8.23×10

-4

lgq e

lgCe

(d)

Figure 1. Adsorption characteristics of furfural onto bamboo charcoal. (a) Effect of initial concentrations on the adsorption capacityof furfural, T = 298 K, m = 75 g/L; (b) effect of temperature on the adsorption capacity of furfural, C = 10 g/L, m = 75 g/L; (c) test ofthe kinetic models defined by Eqs. (1)–(5) for the homogeneous particle diffusion and shrinking core models. m = 75 g/L, T = 298K, C = 10 g/L; (d) linear Langmuir and Freundlich plots for the adsorption isotherm of furfural.

ENERGY SOURCES, PART A: RECOVERY, UTILIZATION, AND ENVIRONMENTAL EFFECTS 1523

Table1.

Parametersof

kinetic

mod

elsof

furfuraladsorptio

non

tothebambo

ocharcoal.

Pseudo

-first-ordermod

elPseudo

-secon

d-ordermod

el

Adsorptio

ncond

ition

q e,exp

(mg/g)

q e,cal(m

g/g)

k 1(m

in−1 )

R2q e

,cal(m

g/g)

k 2(g/(mgmin))

R2

Initialconcentration(g/L)

556.71

41.11

2.42E-03

0.9671

2.44E-04

0.9946

1076.57

59.30

2.08E-03

0.9431

1.09E-04

0.9937

2098.01

74.85

2.35E-03

0.9755

1.21E-04

0.9909

Temperature

(K)

298

76.57

59.30

2.08E-03

0.9431

1.09E-04

0.9937

318

80.72

51.77

2.07E-03

0.9393

1.44E-04

0.9961

328

81.11

45.79

2.20E-03

0.9208

1.89E-04

0.9970

1524 Y. C. LI ET AL.

Shakir and Beheir, (1980) qualitatively demonstrated the adsorption of UO22+ with di-(2-ehtylhexyl)

phosphoric acid (DEHPA) was controlled by united chemical reaction and particle diffusion. Thus,coupling of kinetic mechanisms may play a critical role in this study.

Since the HPDM and SCM models controlled by particle diffusion give better fitting than the othertwo models, the diffusion coefficients for furfural can be calculated by Eqs. (2) and (4) through the slopevalues. The obtained diffusion coefficients are actually a measurement of the mean of the furfuralmolecules related to the adsorption process (Valderrama et al., 2010). The results are also shown inTable 2. For HPDM, the values of diffusion coefficient (De) decrease first, and then increase with theincrease in furfural concentration. However, the diffusion coefficients counted by the SCMdecrease withthe increase of the initial furfural concentration. It may be caused by the accumulation of the productfrom the reacted layer that blocks the progress of furfural within the layer. The diffusion coefficients

0 120 240 360 480 600 7200.0

0.5

1.0

1.5

C = 5 g/L

HPDM: Ln(1-X2)

SCM: 1-3(1-X)2/3+2(1-X)F

(X)

Time/min

0 120 240 360 480 600 7200.0

0.5

1.0

1.5T = 298 K

HPDM: Ln(1-X2)

SPM: 1-3(1-X)2/3+2(1-X)

F(X

)

Time/min

0 120 240 360 480 600 7200.0

0.5

1.0

1.5C = 10 g/L

HPDM: Ln(1-X2)

SCM: 1-3(1-X)2/3+2(1-X)

F(X

)

Time/min

0 120 240 360 480 600 7200.0

0.5

1.0

1.5T = 318 K

HPDM: Ln(1-X2)

SPM: 1-3(1-X)2/3+2(1-X)

F(X

)

Time/min

0 120 240 360 480 600 7200.0

0.5

1.0

1.5

C = 20 g/L

HPDM: Ln(1-X2)

SCM: 1-3(1-X)2/3+2(1-X)

F(X

)

Time/min

0 120 240 360 480 600 7200.0

0.5

1.0

1.5T = 328 K

HPDM: Ln(1-X2)

SPM: 1-3(1-X)2/3+2(1-X)

F(X

)

Time/min

(a) (b)

Figure 2. Particle diffusion models of furfural defined by Eqs. (2) and (4). (a) With different initial concentrations, T = 298 K, m = 75g/L. (b) At different initial temperatures, C = 10 g/L, m = 75 g/L.

ENERGY SOURCES, PART A: RECOVERY, UTILIZATION, AND ENVIRONMENTAL EFFECTS 1525

Table2.

Linear

regression

analysisforHPD

MandSCM.

HPD

MSCM

Cond

ition

F(X)

=f(t)slop

e(m

in−1 )

R2De(m

2 /s)

F(X)

=f(t)slop

e(m

in−1 )

R2De(m

2 /s)

C=5g/L

2.22E-03

0.99

8.21E-12

7.83E-04

0.98

7.50E-11

C=10

g/L

1.89E-03

0.99

6.99E-12

6.88E-04

0.99

2.11E-11

C=20

g/L

2.04E-03

0.99

7.54E-12

7.22E-04

0.99

1.41E-11

T=298K

1.89E-03

0.99

6.99E-12

6.88E-04

0.99

2.11E-11

T=318K

1.96E-03

0.95

7.25E-12

7.11E-04

0.93

2.37E-11

T=328K

2.12E-03

0.91

7.84E-12

7.62E-04

0.90

2.61E-11

1526 Y. C. LI ET AL.

calculated by both models increase slightly with increasing temperatures, indicating that the diffusionresistance decreases with the increase of temperatures. Furthermore, the diffusion coefficient predictedby the HPDM was in the order of 10−12 m2/s, that agrees very well with previous studies that obtainedfrom the resins (Benamor et al., 2008; Cortina andMiralles, 1997), but it was lower than that of the SCM,which had an order of 10−11 m2/s. Benamor et al. also found the diffusion coefficients of HPDMpredicted from cadmium were in the order of 10−12 m2/s, smaller than those of SCM, which had anorder of 10−7 m2/s. The diffusion coefficients predicted by HPDM are lower than those of SCM, whichmight be attributed to the existence of chemical reaction between the furfural and the functional groupson the bamboo charcoal surface (Saha et al., 2004).

3.3. Adsorption isotherm

The adsorption data of furfural were examined by the Langmuir and Freundlich equations and theresults are shown in Figure 1d. As can be seen, the adsorption data can be well described by both theLangmuir and Freundlich isotherm equations. Based on the value of correlation coefficient (R2), theFreundlich isotherm provides a better fitting. It indicated that the adsorption feature of bamboocharcoal takes place by a nonideal adsorption on the heterogeneous surface. In addition, the value of1/n (0.804) is less than 1, which demonstrates that the adsorption condition is favorable.

3.4. Adsorption thermodynamic

To fully recognize the nature adsorption process of bamboo charcoal, a thermodynamic study wasperformed. The Gibbs free energy can be calculated by the Van’t Hoff equation: ΔG° = −RT lnKD,where ΔG° is the change of Gibbs free energy (kJ/mol). KD is defined as a linear or single pointadsorption distribution coefficient, and it is equal to qe/Ce (Sahu et al., 2008). The change in Gibbsfree energy is still associated with the heat of adsorption and entropy change at constant temperatureby ΔG° = ΔH° − TΔS° (Sahu et al., 2008), where ΔH° and ΔS° are the change of enthalpy (kJ/mol)and entropy (J/K), respectively. Combining the two equations above gives lnKD = −ΔG°/(RT) = ΔS°/R − ΔH°/(RT). The value of ΔH° can be obtained from the slope of the equation plot, and theintercept gives the value of ΔS°.

As mentioned in Section 3.1, Ho’s model could fit well the adsorption process. Actually, the rateconstants (k2) of Ho’s model can be used to calculate the activation energy of adsorption from theArrhenius equation (Li et al., 2010): lnk2 = lnA− Ea/(RT).A andEa are theArrhenius factor and activationenergy, respectively. The activation energy could be calculated from the slope of lnk2 versus 1/T.

All of the thermodynamic parameters calculated from the above equations are listed as follows:the KD values are 20.37 L/kg, 23.24 L/kg, and 24.21 L/kg at 298 K, 318 K, and 328 K, respectively.However for ΔG°, it is −7.47 kJ/mol, −8.32 kJ/mol, and −8.69 kJ/mol, respectively. The ΔH°, ΔS°, andEa values are 4.77 kJ/mol, 0.44 J/K, and 2.02 kJ/mol, respectively. Generally, the free energy change isfrom −20 to 0 kJ/mol for physical adsorption, while it is from −400 to −80 kJ/mol for chemicaladsorption (Qiu et al., 2012). From the value of ΔG°, it can be seen that the adsorption of furfural onbamboo charcoal is negative, and is in the range of −20 ~ 0 kJ/mol, which is consistent with that ofother studies (Li et al., 2010; Qiu et al., 2012). It implies the adsorption of furfural is a spontaneous

Table 3. Comparison of adsorption capacities of different adsorbents for furfural.

Adsorbent Adsorption capacity (mg/g) Temperature (K) Reference

Bamboo charcoal 105.15 298 This workPolymeric resin: XAD-4 74.55 303 (Gupta et al., 2001)Bagasse Fly Ash 81.97 293 (Sahu et al., 2007)Activated Carbon: Coconut-based 55 293 (Sahu et al., 2008)Activated Carbon: supplied by Unicarbon, Italian 374 303 (Sulaymon & Ahmed, 2007)Nano-porous adsorbent 196.1 308 (Anbia & Mohammadi, 2009)

ENERGY SOURCES, PART A: RECOVERY, UTILIZATION, AND ENVIRONMENTAL EFFECTS 1527

process of physical adsorption. Moreover, when the temperature increases from 298 K to 328 K, ΔG°changes from −7.47 to −8.69 kJ/mol; thus it is suggested the adsorption is even more spontaneous athigher temperature and similar results can also be found in the literature (Li et al., 2010). The valueof ΔH° provides a measurement of the bonding strength between the adsorbent and adsorbatesurface (Suzuki and Fujii, 2004). The ΔH° with a positive value confirms the endothermic propertyof the overall adsorption process, and it is supported by the phenomenon of higher temperatureresulting in higher adsorption capacity of furfural. Moreover, the positive value of ΔS° indicates therandomness increased at the solid–liquid interface and the degree of freedom of the absorbed specieswas also increased in the adsorption process. In addition, the activation energy (Ea) is 2.02 kJ/mol inthis study, which further confirms that the adsorption process is mainly physical. However, furtherincrease of temperature is harmful for furfural removal, as previously discovered (Li et al., 2013).

3.5. Comparison of different adsorbents

Table 3 lists a comparison of furfural adsorption capacity by different types of adsorbents. Owing tothe variety of parameters and conditions, direct comparison was impossible. However, from Table 3,the adsorption capacity (qm, calculated by the Langmuir isotherm equation) of furfural onto thebamboo charcoal was comparable to or much better than most of the other adsorbents reported inthe literature, which further revealed that bamboo charcoal is feasible to be used as an adsorbent forfurfural removal.

4. Conclusions

Owing to the challenges of the fermentation inhibitor removal for bioethanol production, thebamboo charcoal derived from the pyrolysis process was introduced for this application. Thekinetics, isotherm, and thermodynamic characterizations on the adsorption of furfural were inves-tigated. The adsorption process of furfural onto bamboo charcoal obeys Ho’s pseudo-second-orderkinetic model. Both HPDM and SCM can be used for the description of furfural removal ontobamboo charcoal. The particle diffusion of the HPDM and SCM mechanisms shows a goodapproach to the adsorption data of furfural onto bamboo charcoal, indicating the rate-limitingstep is the adsorbent phase diffusion. Thus, increasing the force of interparticle mass transfer willbe helpful for furfural removal. The furfural diffusion coefficient predicted by HPDM was in theorder of 10−12 m2/s, similar to that reported in the literature, but it was smaller than that calculatedby SCM, which had an order of 10−11 m2/s. The Freundlich equation provided a better fitting to theisotherm data compared to the Langmuir equation, implying that the adsorption feature of thebamboo charcoal takes place by a nonideal adsorption on heterogeneous surfaces. Thermodynamicanalysis suggested ΔH° >0, ΔG° < 0, ΔS° > 0, and Ea = 2.02 kJ/mol, indicating the adsorption offurfural onto bamboo charcoal is a spontaneous physical and endothermic process. Compared withother adsorbents, bamboo charcoal proved to be a promising sorbent for furfural removal. Thispreliminary research may invite further investigation on the property improvement of bamboocharcoal for its utilization in the removal of fermentation inhibitors by the adsorption method.

ORCID

J. A. Shao http://orcid.org/0000-0002-2860-0708

Funding

The authors express their great gratitude for the financial supports from the National Natural Science Foundation ofChina (51376075, 51506071), the Special Fund for Agro-scientific Research in the Public Interest (201303095), and theFundamental Research Funds for the Central Universities.

1528 Y. C. LI ET AL.

References

Allen, S. J., Koumanova, B., Kircheva, Z., and Nenkova, S. 2005. Adsorption of 2-nitrophenol by technical hydrolysislignin: Kinetics, mass transfer, and equilibrium studies. Ind. Eng. Chem. Res. 44:2281–2287.

Anbia, M., and Mohammadi, N. 2009. A nanoporous adsorbent for removal of furfural from aqueous solutions.Desalination 249:150–153.

Balat, M. 2009. Bioethanol as a vehicular fuel: a critical review. Energ. Source A 31:1242–1255.Benamor, M., Bouariche, Z., Belaid, T., and Draa, M. 2008. Kinetic studies on cadmium ions by Amberlite XAD7

impregnated resins containing di (2-ethylhexyl) phosphoric acid as extractant. Sep. Purif. Technol. 59:74–84.Cortina, J., and Miralles, N. 1997. Kinetic studies on heavy metal ions removal by impregnated resins containing di-(2,

4, 4-trymethylpentyl) phosphinic acid. Solvent. Extr. Ion. Exc 15:1067–1083.Dana, P. R., and Wheelock, T. D. 1974. Kinetics of a moving boundary ion-exchange process. Ind. Eng. Chem. Fund.

13:20–26.Demirbas, M. F. 2006. Current technologies for biomass conversion into chemicals and fuels. Energ. Source A

28:1181–1188.Gupta, P., Nanoti, A., Garg, M., and Goswami, A. 2001. The removal of furfural from water by adsorption with

polymeric resins. Separ. Sci. Technol. 36:2835–2844.Hameed, B., and Rahman, A. 2008. Removal of phenol from aqueous solutions by adsorption onto activated carbon

prepared from biomass material. J. Hazard. Mater. 160:576–581.Li, L., Liu, S., and Zhu, T. 2010. Application of activated carbon derived from scrap tires for adsorption of Rhodamine

B. J. Environ. Sci. 22:1273–1280.Li, L., and Zhang, H. 2004. Preparing levoglucosan derived from waste material by pyrolysis. Energ. Source A

26:1053–1059.Li, Y., Shao, J., Wang, X., Yang, H., Chen, Y., Deng, Y., Zhang, S., and Chen, H. 2013. Upgrading of bio-oil: Removal

of the fermentation inhibitor (furfural) from the model compounds of bio-oil using pyrolytic char. Energy Fuel27:5975–5981.

Parawira, W., and Tekere, M. 2011. Biotechnological strategies to overcome inhibitors in lignocellulose hydrolysatesfor ethanol production: review. Crit. Rev. Biotechnol. 31:20–31.

Pehlivan, E., and Arslan, G. 2006. Uptake of metal ions on humic acids. Energ. Source A 28:1099–1112.Qiu, T., Zeng, Y., Ye, C., and Tian, H. 2012. Adsorption thermodynamics and kinetics of p-Xylene on activated

carbon. J. Chem. Eng. Data 57:1551–1556.Ranjan, R., Thust, S., Gounaris, C. E., Woo, M., Floudas, C. A., Keitz, M. V., Valentas, K. J., Wei, J., and Tsapatsis, M.

2009. Adsorption of fermentation inhibitors from lignocellulosic biomass hydrolyzates for improved ethanol yieldand value-added product recovery. Micropor. Mesopor. Mater. 122:143–148.

Saha, B., Gill, R., Bailey, D., Kabay, N., and Arda, M. 2004. Sorption of Cr (VI) from aqueous solution by AmberliteXAD-7 resin impregnated with Aliquat 336. React. Funct. Polym. 60:223–244.

Sahu, A. K., Mall, I. D., and Srivastava, V. C. 2007. Studies on the adsorption of furfural from aqueous solution ontolow-cost bagasse fly ash. Chem. Eng. Commun. 195:316–335.

Sahu, A. K., Srivastava, V. C., Mall, I. D., and Lataye, D. H. 2008. Adsorption of furfural from aqueous solution ontoactivated carbon: Kinetic, equilibrium and thermodynamic study. Separ. Sci. Technol. 43:1239–1259.

Serarols, J., Poch, J., and Villaescusa, I. 2001. Determination of the effective diffusion coefficient of Zn (II) on amacroporous resin XAD-2 impregnated with di-2-ethylhexyl phosphoric acid (DEHPA): Influence of metalconcentration and particle size. React. Funct. Polym. 48:53–63.

Shakir, K., and Beheir, S. G. 1980. Gel-liquid extraction and separation of U (VI), Th (IV), Ce (lll), and Co (ll). Separ.Sci. Technol. 15:1445–1458.

Sulaymon, A. H., and Ahmed, K. W. 2007. Competitive adsorption of furfural and phenolic compounds onto activatedcarbon in fixed bed column. Environ. Sci. Technol. 42:392–397.

Suzuki, M., and Fujii, T. 2004. Concentration dependence of surface diffusion coefficient of propionic acid in activatedcarbon particles. Aiche J. 28:380–385.

Valderrama, C., Barios, J. I., Caetano, M., Farran, A., and Cortina, J. L. 2010. Kinetic evaluation of phenol/anilinemixtures adsorption from aqueous solutions onto activated carbon and hypercrosslinked polymeric resin (MN200).React. Funct. Polym. 70:142–150.

Valderrama, C., Cortina, J., Farran, A., Gamisans, X., and De las Heras, F. 2008. Kinetic study of acid red “dye”removal by activated carbon and hyper-cross-linked polymeric sorbents Macronet Hypersol MN200 and MN300.React. Funct. Polym. 68:718–731.

Xiros, C., Vafiadi, C., Paschos, T., and Christakopoulos, P. 2011. Toxicity tolerance of Fusarium oxysporum towardsinhibitory compounds formed during pretreatment of lignocellulosic materials. J. Chem. Technol. Biotechnol.86:223–230.

ENERGY SOURCES, PART A: RECOVERY, UTILIZATION, AND ENVIRONMENTAL EFFECTS 1529